remove-ai-watermarks

mirror of https://github.com/wiltodelta/remove-ai-watermarks.git synced 2026-07-04 23:47:49 +02:00

Files

T

Victor Kuznetsov 2d5b26ed18 test(eval): vision-transcribed ground truth for qwen_in + clean text-CER numbers

data/qwen_in/ground_truth.json is transcribed by vision (PaddleOCR mangled the
stylized Cyrillic), so the text metric scores variants against an accurate
reference instead of noisy OCR-vs-OCR. Re-measured text CER (controlnet vs qwen)
with this ground truth confirms qwen wins text across EN/RU/ZH: openai_1 0.385 vs
0.241, openai_2 0.341 vs 0.290, gemini_1 (ZH) 0.037 vs 0.000 (perfect Chinese even
at the higher 0.30 strength). Faces still favor controlnet. Refresh the numbers in
docs/known-limitations.md to this cleaner methodology.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

2026-06-20 14:26:23 -07:00

controlnet-removal-pipeline-research.md

docs(restore): document that restore methods REGENERATE, not preserve

2026-06-08 21:08:11 -07:00

known-limitations.md

test(eval): vision-transcribed ground truth for qwen_in + clean text-CER numbers

2026-06-20 14:26:23 -07:00

module-internals.md

feat(scripts): fidelity_metrics.py + correct the qwen-vs-controlnet claim

2026-06-20 09:58:22 -07:00

release-and-distribution.md

Slim CLAUDE.md: move module internals, limitations, landscape research to docs